We propose to test the feasibility of developing an inverted SQUID (Superconducting Quantum Interference Device) microscope for neuroscience research. It is similar to an inverted optical microscope except the objective is replaced by an array of low-temperature dc SQUID-based magnetic field pickup coils (<1 mm diameter) located <50 mu/m below a ultra-thin sapphire window (<100 pm)in a microscope stage. They will be kept at superconducting temperature by liquid helium stored in a reservoir within the microscope body. We anticipate that the sensitivity will be sufficiently high for measuring magnetic fields from as few as 1-10 neurons in brain slices or tissue culture. The existing biomagnetic sensors can measure magnetic fields from hippocampal slices without averaging when neurons in the entire slice are synchronously active, but it is not sensitive enough to measure the magnetic field under normal physiological conditions without blocking inhibitory pathways. The proposed microscope will be useful for studying such activity not only from hippocampal slices, but also from other tissues such as neocortical slices. It will be also useful for estimating the distribution of intracellular currents in excitable tissues because a magnetic field distribution above a thin tissue can be uniquely converted to the current distribution in the tissue. The "current image" may provide a new means to study functions of the cortical neurons. Although still poorly explored, the microscope could be also used to study cellular activities such as phagocytosis or molecular binding (e.g. antigen-antibody binding) by measuring movements and relaxation properties of molecules tagged with magnetic particles such as magnetite or magnetic beads. Recent developments in biomagnetic instrumentation and experimental applications suggest that the construction of such a microscope is feasible. In phase I, we wilt determine (1) noise level of a single magnetic field sensor system as a function of pickup coil diameter to evaluate whether we can build a miniature pickup coil with a noise of about 50 fTHHz needed to achieve the predicted level of sensitivity and (2) minimum thickness of the sapphire window as a function of window diameter that provides sufficient protection from the atmosphere pressure since the effective sensitivity critically depends on the distance between pickup coils and the sample. Once these key features are evaluated, we will construct a multichannel system in phase II.